There are multiple consensus GATA regulatory motifs within the first 2 kilobases of 5' flanking sequences of the mouse, pig, and human SRY genes (Fig. 1A). Although numerous GATA binding sites are present in the different SRY promoters, they are not necessarily species conserved which is to be expected as SRY 5' flanking sequences are generally poorly conserved between mammals 52. As shown in Fig. 1B, we tested the ability of GATA4, at varying doses, to transactivate the SRY promoter from these three species by transient transfection assays in the heterologous HeLa epithelial cell line. HeLa cells do not express GATA4 making it a convenient cell line model for promoter regulation studies involving this factor. Of the three species tested, the pig SRY promoter was the most responsive to GATA4 with an activation of about 3-fold (Fig. 1B, middle panel). This was followed by the mouse Sry promoter which was activated by 2-fold in the presence of GATA4 (Fig. 1B, left panel). This 2-fold activation approached significance but was not significant (P > 0.05). Interestingly, the human SRY promoter was not activated by GATA4 (Fig. 1B, right panel), and this despite the presence of multiple consensus GATA binding motifs and the fact that GATA4 could be strongly overexpressed in HeLa cells (Fig. 1C). Similar results were obtained in other heterologous cell lines such as CV-1 fibroblasts (data not shown). The absence or relatively weak ability of GATA4 to activate the SRY promoter suggested that the factor might not bind to SRY GATA motifs with high affinity. However, EMSA analysis revealed the contrary. As shown in Fig. 2, GATA4 was able to strongly bind to all four GATA motifs of the proximal pig SRY promoter. Similar results were obtained with the GATA binding elements of the mouse and human SRY promoters (data not shown). Thus, the mechanism of GATA4 in SRY gene regulation is unlikely one of direct transcriptional activation by GATA4 alone.

Although SRY 5' flanking sequences are known to be divergent in mammals 2152, in addition to GATA regulatory motifs, 2 partially conserved WT1 binding sites are present within the first 2 kb of 5' flanking sequences of the mouse, pig, and human SRY genes (Fig. 3A). The sequences of the putative WT1 binding sites are shown in Fig. 3B. The TCC repeat elements common to the mouse, pig, and human SRY sequences have been shown to be recognized by both the -KTS and +KTS isoforms of WT1 53. The only exception is the human WT1 binding site 1 (5'-GAGGGGGTG-3') which appears to be preferentially recognized by WT1(-KTS) and not WT1(+KTS) 19. The WT1 isoforms, much like GATA4, were capable of activating the SRY promoter on their own; the magnitude of the activations, however, was greater than GATA4 being on the order of 10–20 fold (Fig. 3D, open bars). Since the WT1 and GATA binding elements lie in close proximity to one another, we tested whether GATA4 could transcriptionally cooperate with WT1 on the SRY promoter by performing co-transfection experiments in heterologous HeLa cells (Fig. 3D, black bars). The three species-specific SRY promoter constructs used in the co-transfection experiments contained both WT1 binding sites and 4 or more consensus GATA elements. On the mouse Sry promoter, both WT1 isoforms strongly synergized with GATA4 with activations reaching 30-fold (Fig. 3D, left panel). Interestingly, the level of GATA4/WT1 synergism was higher with WT1(+KTS) than with WT1(-KTS). Similar results were also observed on the pig SRY promoter where activations reached nearly 50-fold (Fig. 3D, middle panel), as well as in other cell lines such as TM4 mouse Sertoli cells (data not shown). In contrast to the mouse and pig, the human SRY promoter was strongly activated by WT1(-KTS) alone and synergism with GATA4 only occurred with the WT1(+KTS) isoform (Fig. 3D, right panel). Western blot analysis confirmed the heterologous nature of HeLa cells since they are null for both GATA4 and WT1 (Fig. 3C). It also showed that the GATA4/WT1 synergism was not due to variable expression of either GATA4 or the WT1 isoforms since all proteins (whether expressed individually or in combination) were present at similar levels when overexpressed in HeLa cells (Fig. 3C).

To assess the binding site requirements for transcriptional synergism between GATA4 and WT1, mouse Sry promoter deletion and mutation constructs were prepared and used in co-transfection assays (Fig. 4A). Interestingly, the -340 bp Sry promoter construct (containing an intact proximal WT1 site and no consensus GATA sites) was synergistically activated by GATA4 and WT1 to the same extent as the full-length (-1090 bp) Sry construct. This suggests that GATA4 binding to its consensus sites on the Sry promoter is dispensable for maximal synergism with WT1. Deletion or mutation of the proximal WT1 site, however, abolished the activation by WT1 alone and markedly decreased the synergism between GATA4 and WT1 (Fig. 4A). Thus, in contrast to GATA4, binding of WT1 to its consensus element is critical for GATA4/WT1 synergism on the mouse Sry promoter. Interestingly, GATA4/WT1 synergism was not completely abrogated with the WT1 mutant constructs. The remaining synergism is likely due to two low affinity (not perfect consensus) GATA binding motifs present within the -340 to -70 bp sequence since a further deletion to -40 bp eliminated all WT1/GATA4 synergism (Fig. 4A). These two GATA motifs, named sites A (GTATCT) and B (GTATCT), are unique to the mouse Sry promoter sequence. Although these sites only weakly bind GATA4 protein (Fig. 4B), they are nonetheless functional as revealed by transfection assay (Fig. 4C). The remaining GATA4/WT1 synergism observed on the -340 bp Sry construct harboring the WT1 site mutation (-340 bp WT1 mut.) was abolished when a truncated wild-type GATA4 protein (aa 201–440) was substituted with two GATA4 mutants (C294A or ΔT279) that we have previously shown to be unable to bind to DNA 54. Thus, GATA4 binding to these two low affinity GATA motifs likely contributes to the observed GATA4/WT1 synergism on the mouse Sry promoter. We still cannot rule out the possibility, however, of additional unknown regulatory elements (present between -340 and -70 bp) that might be indirectly activated by GATA4/WT1 overexpression in our heterologous HeLa cell line model.

In a DNA-binding experiment (Fig. 5A), WT1 (+ or - KTS) binding to a known consensus WT1 probe was competed using either unlabeled probe or an oligonucleotide corresponding to the proximal WT1 element (site 1) of the mouse Sry promoter. Thus, these results demonstrate that the proximal WT1 site (site 1) is efficiently bound by both WT1 isoforms. The functional importance of the proximal WT1 binding site was confirmed using homologous (WT1 and GATA4-expressing) PGR 9E11 cells, a pig genital ridge cell line 21. As shown in Fig. 5B, PGR cells express abundant WT1 and GATA4 proteins. Transfection studies carried out in this cell line clearly showed that the deletion or mutation of the proximal WT1 element reduced Sry promoter activity to the same level as the minimal promoter (approximately 33% of the intact construct; Fig. 5C). Thus, WT1 binding to its proximal site is crucial for full basal activity of the Sry promoter. In our Sry promoter deletion experiments (Fig. 4A), we found that GATA binding to its consensus motifs was dispensable for maximal GATA4/WT1 synergism suggesting that GATA4 cooperates with DNA-bound WT1. As shown in Fig. 5D, a ChIP assay using pig PGR cells confirmed that GATA4 is indeed associated with the endogenous SRY promoter in close proximity to the proximal WT1 binding site (ChIP targeting region 2). In contrast, a more distal SRY promoter fragment lacking the critical WT1 site (ChIP targeting region 1) could not be amplified.

The strong transcriptional cooperation between GATA4 and WT1 on the SRY promoter suggests that both factors contact each other through a direct protein-protein interaction. The essential nature of this interaction is all the more evident given that GATA4 binding to its consensus motifs is not required for maximal synergism with WT1 (Fig. 4A). To map the domains of the GATA4 protein that interact with WT1 and are important for synergism on the Sry promoter, a series of truncated GATA4 proteins were constructed as depicted in Fig. 6A. The GATA4 protein contains two independent activation domains that flank its DNA-binding domain. Deletion of the N-terminal region (aa 1–200) had no significant effect on the transcriptional synergism between GATA4 and WT1(+KTS) (Fig. 6A). However, GATA4/WT1 synergism was lost using proteins consisting solely of the N-terminal region (aa 1–133), zinc finger region (aa 201–322) or C-terminal domain (aa 302–440). Thus, the zinc finger and C-terminal domains of the GATA4 protein are both required for its ability to transcriptionally cooperate with WT1. Next, the domain of the GATA4 protein involved in the direct physical interaction with WT1 was mapped using in vitro pull-down assays (Fig. 6B). Consistent with the transcription data (Fig. 6A), the WT1 protein (- or +KTS) could physically interact with the intact (wild-type) GATA4 protein (Fig. 6B, far left panel) but not with mutant GATA4 proteins that removed the zinc fingers or C-terminal portions of the protein (Fig. 6B, 3 rightmost panels). Thus, the zinc finger and C-terminal domains of GATA4 are also essential for the physical interaction with WT1. The ability of GATA4 to interact with WT1 in the absence of DNA-binding was further confirmed by a co-immunoprecipitation experiment (Fig. 6C).

Since we found in the present study that GATA4 and WT1 could cooperate to regulate transcription from the SRY promoter, we surmised that the expression of other gonadal genes might also be modulated by this functional cooperation. Another obvious target in the sex determination/sex differentiation pathway is the AMH gene since its proximal promoter region contains species-conserved binding elements for both GATA4 and WT1 [see Additional file 1]. Both factors are known regulators of AMH transcription 4548, which is mediated through their direct association with the proximal AMH promoter as shown by ChIP (ref. 45 and data not shown). To better define the role of GATA4 as a transcriptional partner for WT1 in the cell-specific and developmental regulation of the AMH gene, we tested whether GATA4 and WT1 could synergistically activate the AMH promoter. In our co-transfection experiments, we used the mouse -180 bp Amh promoter which contains the GATA and WT1 binding sites in their normal context, as well a series of synthetic constructs containing either GATA and/or WT1 elements placed upstream of the minimal Amh promoter (Fig. 7). Although crucial for Amh transcription in vivo 50, GATA4 by itself is a poor activator of the Amh promoter (Fig. 7A). While WT1 alone weakly activated the Amh promoter, and in contrast to the Sry promoter, only the -KTS isoform was transcriptionally active. In the presence of both proteins, however, a strong synergistic activation was observed not only on the native -180 bp Amh promoter but also a synthetic reporter consisting of two consensus GATA motifs placed upstream of the minimal (-65 bp) Amh promoter containing an intact WT1 binding element (Fig. 7A and 7C). Finally, to assess the binding site requirements for GATA4/WT1 synergism, mutations or deletions of the Amh promoter were generated (Fig. 7B and 7D). Although activation by GATA4 alone was retained using a construct with intact GATA motifs but a mutated WT1 element, GATA4/WT1 synergism was lost (Fig. 7B). Similarly, no GATA4/WT1 synergism was observed on a construct containing an intact WT1 binding site but deleted of the GATA motifs (Fig. 7D). Thus, again unlike the SRY promoter, maximal GATA4/WT1 synergism on the AMH promoter requires both GATA4 and WT1 binding.

There are several naturally occurring WT1 mutants known to cause male pseudohermaphroditism with retention of Müllerian ducts (reviewed in 55). One of these mutants, WT1 R394W, was shown to be a poor activator of the AMH promoter 45. To better explain the molecular mechanism behind the human phenotype associated with the mutation, we were interested in verifying whether the WT1 R394W mutant could still transcriptionally cooperate with GATA4. As shown in Fig. 8, the WT1 R394W mutant, unlike its wild-type counterpart (- or + WT1 isoforms), not only failed to activate the Amh promoter but also failed to synergize with GATA4. This lack of synergism was not related to the stability or expression level of the mutant protein (data not shown). Rather, the lack of synergism is most likely attributable to a decrease in DNA binding affinity of the mutant WT1 protein for its binding element 45. This result therefore supports our previous observation that GATA4/WT1 synergism on the AMH promoter absolutely requires binding of both factors to their respective regulatory elements (Fig. 7). Thus, a failed transcriptional synergism between GATA4 and the mutant WT1 protein at the level of the AMH promoter could be a contributing factor to the improper male sex differentiation seen in WT1 R394W patients involving insufficient AMH production.

The mouse, pig and human SRY promoters were amplified by PCR using the corresponding species-specific genomic DNA as template and the following pairs of primers: mouse (forward: 5'-CGGGATCCGCTGTATTGTCAATAAAACAG-3', reverse: 5'-GGGGTACCGACAATTGTCACCAGTCCC-3'); pig (forward: 5'-CGGGATCCTTTGAGTTCCAAGG-3', reverse: 5'-GGGGTACCGAAAAGGGGGAGGAAGCG-3'); human (forward: 5'-CGGGATCCAATTCATATAGCTTTTTGTGTCC-3', reverse: 5'-GGGGTACCTCAACACCCCCTCAAC-3'). The different promoters fragments were subsequently cloned into the BamHI/KpnI site of a modified luciferase expression vector 65. Deletion constructs for the mouse Sry promoter were generated by PCR using the following pairs of primers: -820 bp (forward: 5'-CGGGATCCGCCGTAGTAGACTATGATAC-3', reverse: 5'-GGGGTACCGACAATTGTCACCAGTCCC-3'); -340 bp (forward: 5'-CGGGATCCCTTTCCACTACTTTTGCA-3', reverse: 5'-GGGGTACCGACAATTGTCACCAGTCCC-3'); -40 bp (forward: 5'-CGGGATCCTTACACACGTTAAATATTAAAATC-3', reverse: 5'-GGGGTACCGACAATTGTCACCAGTCCC-3'). Promoter fragments were then cloned into the luciferase reporter as described above. The -340 mouse Sry promoter construct with a specific deletion of the WT1 response element (ΔWT1) was achieved by cloning a -340 to -70 bp Sry promoter fragment, obtained by PCR (forward primer: 5'-CGGGATCCCTTTCCACTACTTTTGCA-3', reverse primer: 5'-GAAGATCTCTAGTCCAGCCCAACTAATC-3'), in front of the minimal mouse Sry promoter construct. A -340 bp mouse Sry promoter construct harboring a mutation that inactivates the WT1 element (GGAGGAGGGA to GGTGTTGGGT) was generated using the QuikChange XL mutagenesis kit (Stratagene, La Jolla, CA) according to the manufacture's recommendations along with the following oligonucleotides – sense: 5'-GACTAGGGAGGTCCTGAAGGTGTTGGGTTAAATATTTTCTTACAC-3'; antisense: 5'-GTGTAAGAAAATATTTAACCCAACACCTTCAGGACCTCCCTAGTC-3'. The natural and synthetic murine Amh-luciferase promoter constructs have been described previously 31. The synthetic Amh promoter construct containing a WT1 binding site mutation was generated by site-directed mutagenesis using the following primer pair – sense: 5'-TACAGCAAGGCCCGGGCGGCCCCGCTTATATGTA-3'; antisense: 5'-TACATATAAGCGGGGCCGCCCGGGCCTTGCTGTA-3'. Finally, the WT1(-KTS) DDS (R394W) mutant was also made by site-directed mutagenesis using the wild-type WT1 cDNA as template and the following primers – sense: 5'-CTTCAGATGGTCGGACCAGGAAAACTTTCGCTG -3'; antisense: 5'-CAGCGAAAGTTTTCCTGGTCCGACCATCTGAAG-3'. The GATA4 expression vectors (wild-type, deletions and DNA-binding mutants) have been described previously 31546266. Expression vectors for the wild-type mouse WT1 isoforms were kindly provided by Dr. Jerry Pelletier (McGill University). All plasmid constructs were verified by sequencing.

Pig genital ridge PGR 9E11 cells 21 were grown in DMEM supplemented with 10% newborn calf serum at 37°C and 5% CO2. The human cervical carcinoma HeLa cell line was maintained in 1:1 DMEM/Ham's F-12 containing 10% fetal bovine serum. HeLa cells were transfected in 24-well culture plates using the calcium phosphate precipitation method 67. PGR 9E11 cells were seeded in 12-well plates and transfected using the LipofectAMINE reagent (Invitrogen, Burlington, Canada). In brief, 2.5 μg of SRY promoter-luciferase reporters were transfected with 2.5 μl of LipofectAMINE reagent per well in serum- and antibiotic-free media. A renilla (phRL-TK) luciferase plasmid (Promega, Madison, WI) was used as an internal control. Complete medium was added 5 h after transfection. The next day, the cells were harvested and analyzed using the Dual-Luciferase reporter assay system from Promega. The data reported represent the averages of at least three experiments, each done in duplicate. For the experiments presented herein, the internal control was unaffected by either GATA4 or WT1 overexpression [see Additional file 2].

Nuclear extracts were prepared by the procedure outlined by Schreiber et al. 68. In Western analyses, 10-μg aliquots of nuclear extracts from untreated PGR 9E11 or transfected HeLa cells with GATA4 and/or WT1 expression vectors were separated by SDS-PAGE and then electrotransferred to Hybond PVDF membranes (GE Healthcare Life Sciences, Baie d'Urfé, Canada). Immunodetection of the GATA4 protein was achieved using a GATA4 antiserum (catalog # sc-1237; Santa Cruz Biotechnology, Santa Cruz, CA). WT1 proteins were detected using an anti-WT1 antibody (catalog # sc-192; Santa Cruz) and a VECTASTAIN-ABC-Amp Western blot detection kit (Vector Laboratories Canada, Burlington, Canada). Duolux (Vector) was used as chemiluminescent substrate.

To assess binding of GATA4 to the consensus GATA sites in the pig SRY promoter, in vitro translated GATA4 protein was prepared using the TNT system from Promega. DNA binding assays were performed using a ³²P-labeled double-stranded oligonucleotide corresponding to the proximal GATA site (GATA site 1) of the pig SRY promoter – sense oligo: 5'-GATCCGCCTTATTATCATAATAAA-3'; antisense oligo: 5'-GATCTTTATTATGATAATAAGGCG-3'. Competition with an oligonucleotide containing a mutated GATA motif (GATA to GGTA) was used to confirm the specificity of GATA4 binding – GATA_mut site 1 (sense: 5'-GATCCGCCTTATTACCATAATAAA-3', antisense: 5'-GATCTTTATTATGGTAATAAGGCG-3'). Competition experiments using a series of double-stranded oligonucleotides were then used to confirm GATA4 binding to the more distal GATA motifs in the pig SRY promoter – GATA site 2 (sense: 5'-GATCCATTGGGTTATCTTGAATCA-3', antisense: 5'-GATCTGATTCATGATAACCCAATG-3'); GATA site 3 (sense: 5'-GATCCACATACTGATAATCATCAA-3', antisense: 5'-GATCTTGATGATTATCAGTATGTG-3'); GATA site 4 (sense: 5'-GATCCCCAAGGTTATCTGTTTTTA-3', antisense: 5'-GATCTAAAAACAGATAACCTTGGG-3'). Recombinant GATA4 protein was also used to assess the affinity of GATA4 binding to two non-consensus GATA motifs (named sites A and B) in the proximal mouse Sry promoter. For this specific experiment, the consensus GATA element of the proximal murine Star promoter (sense: 5'-GATCCACTTTTTTATCTCAAGTGA-3'; antisense: 5'-GATCTCACTTGAGATAAAAAAGTG-3') was used as labeled probe. Competition using oligos corresponding to sites A (sense: 5'-GTTCTTTGTATCTTAATACT-3'; antisense: 5'-AGTATTAAGATACAAAGAAC-3') and B (sense: 5'-CTTGACAGTATCTAGGTTCA-3'; antisense: 5'-TGAACCTAGATACTGTCAAG-3') was used to assess the affinity of GATA4 binding to these two sites. A similar approach was used to demonstrate WT1 binding to the proximal WT1 element in the mouse Sry promoter. In vitro translated WT1(-KTS) and WT1(+KTS) proteins were prepared using the TNT system (Promega). WT1 binding assays were performed using a ³²P-labeled double-stranded oligonucleotide corresponding to the distal WT1 element of the human AMH promoter (sense oligo: 5'-GGATCACTGGGGAGGGAGATAGGA-3', antisense oligo: 5'-GATCTCCTATCTCCCTCCCCAGTG-3') which has been described as a consensus binding site for both the -KTS and +KTS isoforms of WT1 20. Competition experiments using a double-stranded oligonucleotide corresponding to the proximal WT1 site in mouse Sry promoter (sense oligo: 5'-GGATCCTGAAGGAGGAGGGATAA-3', antisense oligo: 5'-GATCTTATCCCTCCTCCTTCAGG-3') was used to confirm WT1 binding to the Sry promoter. For all EMSA experiments, binding reactions and electrophoresis conditions were as described previously 48.

Recombinant histidine-tagged WT1 fusion proteins (HIS-WT1) were obtained by cloning the coding region of the mouse WT1 protein in-frame with HIS using the commercially available pRSETB fusion protein vector from Invitrogen. The resulting construct was introduced into the Escherichia coli strain BL21, and the fusion protein was produced by inducing the bacterial culture with IPTG. After induction, the bacterial culture was lysed by sonication and the fusion protein was purified using a TALON metal affinity resin (BD Biosciences, Mississauga, Canada) according to instructions outlined by the manufacturer.

In vitro pull-down (protein-protein interaction) assays were done using ³⁵S-labeled in vitro translated GATA4 proteins (wild-type and deletion mutants) and the purified HIS-WT1 and HIS-LacZ fusion proteins coupled to TALON metal affinity resin. The ³⁵S-labeled GATA4 proteins were obtained using the TNT system as described above; the amino acid positions of the different GATA4 proteins used are given in Fig. 6A. Proteins were incubated in 500 μl binding buffer (20 mM Hepes pH 7.9, 70 mM KCl, 20% Glycerol, 0.5% Triton X-100) supplemented with 0.01% BSA for 30 min at room temperature. Bound immunocomplexes were washed three times in washing buffer (60 mM NaH2PO4, 600 mM NaCl, 20% Glycerol, 0.75% Triton X-100), resuspended in Laemmli buffer, and subjected to SDS-PAGE. Proteins were finally electrotransferred to Hybond PVDF membrane and visualized by autoradiography.

Nuclear extracts were prepared from HeLa cells transfected with expression vectors for GATA4 or a combination of GATA4 and WT1. A 100-μg aliquot of each extract was then immunoprecipitated overnight using the WT1 antibody in binding buffer as described previously 69. Immunocomplexes were then collected by incubation with 20 μl of protein G-Sepharose beads (GE Healthcare Life Sciences) at 4°C for 1 h. Bound immunocomplexes were washed four times with binding buffer, resuspended in 30 μl of 1× Laemmli buffer, and subjected to SDS-PAGE. Proteins were electrotransferred to PVDF membrane and subjected to immunoblotting using a 1:1000 dilution of the GATA4 antibody and detected with the Vectastain-ABC-Amp Western blot detection kit.

PGR cells (4 × 10⁷) were cross-linked with 1% formaldehyde (final concentration) for 10 min at 37°C. The cells were then washed 3 times with PBS, resuspended in ChIP lysis buffer [1% SDS, 10 mM EDTA, and 50 mM Tris-HCl (pH 8.0)], and incubated for 10 min at 4°C. The lysed cells were sonicated on ice using a BRANSON 450 Sonicator for 5 cycles of 30-sec pulses at 3.5 output control and 80% duty cycle in order to obtain DNA fragments between 200 bp and 1000 bp in size. After centrifugation of the solution for 10 min at 4°C, the supernatant was precleared for 2 h at 4°C with 20 μl of a protein G-Sepharose bead slurry (GE Healthcare Life Sciences) that had been prepared by washing it 3 times with RIPA buffer [50 mM Tris-HCl (pH 8.0), 5 mM EDTA, 150 mM NaCl, 0.1% SDS, 0.5% Igepal CA-630 (Sigma-Aldrich Canada, Oakville Canada)]. After centrifugation for 10 min at 4°C, 500 μl of supernatant was subjected to overnight incubation at 4°C with 0.4 μg anti-GATA4 or control goat IgG (Santa Cruz) in immunoprecipitation buffer [20 mM Tris-HCl (pH 8.0), 2 mM EDTA, 150 mM NaCl, 1% Triton X-100]. The immunocomplexes were recovered by a 2 h incubation with 20 μl protein G Sepharose beads. The precipitates were washed 5 times with 1 ml RIPA buffer, 5 times with 1 ml high salt RIPA buffer [50 mM Tris-HCl (pH 8.0), 5 mM EDTA, 500 mM NaCl, 0.1% SDS, 0.5% Igepal CA-630], 3 times with LiCl buffer (250 mM LiCl and 0.1% Igepal CA-630), and finally twice with 1 ml of TE buffer [10 mM Tris-HCl (pH 8.0), 1 mM EDTA]. Protein-DNA complexes were eluted from protein G Sepharose beads by addition of elution buffer (1% SDS, 0.1 M NaHCO₃) and rotation at room temperature for 15 min. Cross-links were reversed by addition of 200 mM of NaCl and heating at 65°C for 4 h. Proteins were degraded using proteinase K treatment (1 h at 55°C) and the DNA fragments were purified using a QIAquick gel extraction kit (Qiagen, Mississauga, Canada). PCR amplifications were done using 1 μl of input chromatin sample and 5 μl of GATA4-immunoprecipitated DNA sample with primers specific for either the distal (ChIP region 1) pig SRY promoter (forward primer: 5'-GTATTCACTTATTTCATTTGGTAAGCCA-3'; reverse primer: 5'-CGAGGTGAACATAACACTTC-3') or the proximal (ChIP region 2) pig SRY promoter (forward primer: 5'-TACTGGGGGCGGAGAAATTG-3'; reverse primer: 5'-GGGTCGCTTGACACGATCCT-3'). PCR amplifications were carried out at 94°C for 5 min, followed by 40 cycles of 94°C for 30 sec, 60°C for 30 sec, and 72°C for 30 sec, and a final extension of 5 min at 72°C. The PCR products were analyzed by electrophoresis on a 2% ethidium bromide-stained agarose gel. ChIP results were confirmed by at least 2 separate experiments.

Statistical comparisons between multiple groups were analyzed by one-way analysis of variance (ANOVA) followed by a Student-Newman-Keuls test. Where normality and/or equal variance among groups was not met, data were analyzed by a Kruskal-Wallis ANOVA followed by a Student-Newman-Keuls test to detect significant differences. P < 0.05 was considered significant. All statistical analyses were done with the aid of the SigmaStat 3.5 software package (Systat Software, Inc., Point Richmond, CA).